Physician & Physician Assistant Attitudes and Referral Habits Concerning Chiropractic
Shawn M. Neff, DC, MAS, FACO1, 2, Regina J. Jordan, MSN, RN3
Staff Chiropractor, Martinsburg Veterans Affairs Medical Center, Martinsburg, WV1
Adjunct Faculty, Palmer College of Chiropractic2
Registered Nurse, Martinsburg Veterans Affairs Medical Center, Martinsburg, WV3
Journal of the Academy of Chiropractic Orthopedists
December 2016, Volume 13, Issue 2
This article is available from: https://ianmmedicine.org © 2016 Neff/ Jordan and the Academy of Chiropractic Orthopedists. This is an Open Access article which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Objectives: This study looks to utilize survey methods to evaluate the factors which may influence the attitudes and referral habits to chiropractors among physician and physician assistants (PAs) within an integrated health care environment where access to chiropractic services is referral-dependent. This study compares the self-reported data on referral habits to actual referral data.
Methods: This study was approved by the Washington DC VAMC institutional review board. Subjects were members of the medical staff of the Martinsburg VA Medical Center. They completed a written informed consent. Data was collected using a tool developed for this study. Data was then placed in numeric categories and analyzed using SPSS 13.0. Statistical analysis consisted of descriptive statistics, bivariate correlation, and factorial analysis of variance.
Results: The analyses were carried out on 26 competed surveys. The mean average knowledge of chiropractic (self- rated) and opinion of chiropractic was 4.8 on a 0-10 scale with a range from 1-9. 42% of subjects referred to chiropractors. The average satisfaction of the referring provider with the results of referral was 8 (0-10 scale). The presence of chiropractor on staff had a positive effect on both opinion and referral habits.
Conclusions: The tool developed does effectively gather the data sought. Having a chiropractor on staff does positively impact the opinions medical providers have concerning chiropractic, as well as the referral habits to chiropractors. Demographically PAs have an overall higher opinion of chiropractic than physicians. The more medical providers know about chiropractic, the higher their opinion. The majority of providers who refer more do so because they feel it is the best treatment in that case. Overall satisfaction with referrals to the chiropractic service is very high (8/10).
With the increasing popularity of chiropractic care in the United States, inter-professional relationships between conventional trained physicians and doctors of chiropractic (DCs) will have an expanding impact on patient care. 1 Which factors are most associated with a positive attitude about chiropractors? Studies have shown that younger doctors are more likely to have positive attitudes concerning chiropractic. 2, 3, 4 Other variation in physicians’ knowledge and referral behavior might be due to historical and political circumstances, ethnic traditions, availability and regional demand.5 Sikand found that female doctors are more likely to discuss or refer for complementary and alternative medicine (CAM).6 No studies were found evaluating differences in attitudes between medical providers who work in hospitals with chiropractors on staff versus those who work in hospitals without staff chiropractors. The Department of Veterans Affairs (VA) is the perfect living laboratory for examining this question. It is a large integrated health care system with a naturally created experimental and control group due to the fact that some VA medical centers have chiropractors on staff, while others do not.
This study served as a pilot of a newly designed questionnaire and protocol to allow for future comparison. This study also examined the level of correlation between different characteristics and positive attitudes and referral habits.
MDs, DOs, and Physicians Assistants (PAs) were surveyed using a tool developed for this study (Appendix A). Questions covered demographic information (age, specialty, time elapsed since training), attitudes about chiropractic and referral habits to chiropractors. The data from the survey was paired with the provider’s actual referral data from the Veterans Health Information Systems and Technology Architecture (VISTA) system. A comparison was made between self-reported and actual data.
An 11 point numeric scale (-5-+5) was used for rating the opinion of chiropractic and chiropractors, as well as the effect that having a chiropractor on staff has on the opinion of or referral habits to chiropractors. A six category, nominal scale, was used for reporting the number of patients seen and the frequency of referrals. From the VISTA system, actual referral data was placed in the same nominal scale.
The questionnaire was reviewed by colleagues and administered to a small group who provided feedback to improve question content and readability. Changes were made to improve the tool based on comments. The small sample showed validity and concurrence between self-report and actual referral data. It was determined that a sample size of 20% of the facilities medical providers would be sufficient for a pilot based on Baker’s findings that 10-20% of the sample size of a study is reasonable for pilot enrollment.9
Statistical analysis was completed using factorial analysis of variance and bivariate correlation. Parallel forms reliability and paired sample T-test were used to test the similarity and difference between the survey referral data and the VISTA referral data. Correlation was used for all factors to isolate effects. Results from the statistical analysis were used to compare the factors influencing referral rates and attitudes, and to determine the relationship between actual referral data and the provider’s self-assessment of referral patterns.
Details of Subjects
For the purposes of this study, the entire population of staff MDs, DOs and PAs of VA Medical Center (VAMC) in Martinsburg, WV were eligible as long as they consented and had been on staff for at least three months prior to surveying. The age range of subjects is between 20 and 80 years of age. Providers who had practiced at VAMC Martinsburg less than three months prior to surveying were excluded.
Subjects were members of the medical staff at VAMC Martinsburg. Informed consent was obtained as well as clearance by the Institutional Review Board (IRB), Research and Development (R&D) committee and the bargaining unit (union) representatives. The study was announced during staff meetings. Staff were also approached privately by department by the principle investigator (PI) or research assistant for participation. In order to minimize selection bias, every provider in a department or service was invited. If willing to participate and sign the informed consent document, they completed the questionnaire. The questionnaires were completed and sealed in an envelope and delivered to the research assistant.
This group was selected for several reasons. The first being convenience of using staff all at one medical center as a pilot and allowing for less variability in work conditions. Second, by using only these three provider types only one union had to approve. Those with less than 3 months of service were excluded because the actual referral data was averaged over 3 months retrospectively to minimize bias or focal anomalies in referral rate. Those with less than 3 months of service could not be averaged in the same manner and were, for that reason, excluded.
In total 26 subjects completed the survey and returned it to the research team (Table 1). There are 116 physicians and physician assistants on staff at VAMC (88 MDs, 5 DOs, and 23 PAs) so this represents 22.4% of the provider population. Of the respondents, 14 were MDs, while 12 were PAs. In addition, eight respondents were primary care providers, 12 were specialists and 2 were emergency department providers. Three were hospitalists and three were compensation and pension doctors, who perform disability examinations. Demographic breakdown on the participants revealed that 14 subjects were male with the remaining 12 being female. Only 22 answered concerning their age, of these five were 20-34, nine were 35-44, five were 45-54, and three were 55-64. Over half of the respondents (57.7%) had practiced for 10 or fewer years. The mean average knowledge of chiropractic (self-rated) was 4.8 on a 0-10 scale with a range from 1-9.
Valid N (listwise)
Table 1: Descriptive Statistics
The opinion of chiropractors had the same mean and range. Almost 70% see 35 or more patients per week. Approximately 80% of these patients were outpatients. 42% of subjects reported referrals to a chiropractor. The average satisfaction with the treatment was 8 on a 0-10 scale (Table 2). The effect of the presence of a staff chiropractor on opinion of chiropractic and on referral habits was rated on a -5 to 5 scale and the mean for the effect on referrals was 2.25 and the mean for effect on opinion was 1.85. The findings were assessed utilizing SPSS 13.0.
Table 2: Satisfaction
Pearson Correlation was significant at the 0.05 level between provider type and opinion about chiropractic/chiropractors as well as between number of patients seen per week and opinion about chiropractic/chiropractors, and the awareness of research on efficacy of chiropractic and opinion about chiropractic/chiropractors. There was significance at the 0.01 level for self-perceived level of knowledge about chiropractic and opinion about chiropractic/chiropractors and level of education of chiropractors and opinion. Significance at the 0.01 level was also observed between the number of consults and the reason for the consult.
Factorial ANOVA revealed no significant effects when using opinion of chiropractic as an independent variable and age, gender, provider type, specialty, knowledge of chiropractic, staff chiropractor, experience, and satisfaction with chiropractic treatment as fixed variables. The same was true when evaluating number of chiropractic consults ordered using the same fixed variables.
Actual referral data was gathered from the VISTA system from 12/01/2010 through 05/30/2011. During that time there were 162 consults placed to the chiropractic service (106 from MDs, 6 from DOs, 19 from PAs, 28 from nurse practitioners (NPs), 1 from a psychologist, and 3 from physical therapists (PTs)). Of these consults the majority (136 or 84%) were from primary care (nine from the Operation Enduring freedom/Operation Iraqi Freedom clinic (6.6%), 24 from the women’s clinic (17.6%). The emergency department was responsible for nine consults (5.6%). Geriatrics and long term care referred four patients (2.5%) and medical and surgical specialists referred nine (5.6%). Physical therapists referred three patients to the chiropractic clinic (1.9%). Bivariate Correlation by paired sample between the self-reported referral data and the actual referral data from the VISTA system was significant at the 0.01 level.
** Correlation is significant at the 0.01 level (2-tailed).
Table 3: Correlation between self-reported referral rates and actual referral rates
There is currently very little data on medical attitudes and referral habits to chiropractors in an integrated medical center setting. Even less is known about the differences in attitudes and referral habits between medical centers with chiropractors on staff and those without. This is especially important in looking at the real access to care in the VA when there are many medical centers without staff chiropractors. This study is the essential groundwork for looking at inter-facility differences.
The data collected shows that there is a relationship between the self-perceived knowledge about chiropractic and the opinion of chiropractic, with those knowing less having a lower opinion. This supports previous studies which show that greater knowledge is associated with acceptance and respect between professions,10 and studies which show that providers with more knowledge of chiropractic have higher opinions of it2. Additionally the knowledge of research on the efficacy of chiropractic for different conditions and the opinion of chiropractic show a significant relationship. The relationship between knowledge and opinion is a positive one and shows that those who are less aware of the research have a lower opinion, or that those with a lower opinion are less aware of the research. This relationship is important because Goldszmidt et al noted that self-reported knowledge of chiropractic was poor: the proportion of general practitioners (GPs) who reported a high level of knowledge about chiropractic was only 10%.11 This lack is significant because patients are influenced by what their physician recommends12 and therefore the physicians’ perception of Chiropractic influences the patients’ perception of Chiropractic. This is further compounded by the fact that patients do not generally share the fact they visit complementary practitioners with their general practitioners (because of the perception that the GP would disapprove). While an important factor influencing general practitioners’ opinions about chiropractic appears to be patients’ experience, which may bias physician opinions and subsequently influence referral habits.5 This creates a feedback loop which reinforces past behavior as future behavior.
There is also a relationship between whether the provider is an MD or a PA and their opinion of chiropractic with PAs having a higher opinion of chiropractic on average than physicians. This supports the findings of Isberner et al that most (66%) of the PAs they surveyed felt that while we need to be cautious in our claims, a number of CAM therapies hold promise for the treatment of symptoms, conditions, and/or diseases.13 Houston et al found a significant relationship between knowledge level and recommendation for CAM among PAs.14
The majority of consults placed were placed by providers who believed it was the best treatment available. Additionally, as mentioned previously those who do refer to chiropractors have a high satisfaction with the results of the consult (mean of 8/10). The providers who see the most patients per week seem to have the lowest opinion of chiropractic, and also report the least knowledge of chiropractic.
Provider’s self-reported knowledge of chiropractic is significantly correlated with the answers to questions about chiropractic education (p=0.001) and research (p=0.034). This shows validation of the self-reported knowledge of chiropractic. Interestingly though, the number of referrals made to chiropractors was not significantly correlated to opinion, knowledge, or even satisfaction with chiropractic. However, with the small sample size it is difficult to generalize results, and in larger studies relationships may appear which are not apparent in this study.
When comparing the self-reported referral data with actual referral data from the VISTA system there is a significant correlation. This will allow future surveys to be anonymous as there will not be the need to collect names in order to pair the self-reported referral data to the actual data. This will decrease the risk to the subjects.
The actual referral data showed that the vast majority of consults do come from primary care providers. This is to be expected because chiropractors in the VA function as specialists, and because of the primary care gatekeeper model the VA follows. The majority of the respondents were not primary care providers. This may have had an effect on some of the results as many of the specialists surveyed had never referred to the chiropractic service. Most noted it was not because of negative feelings about chiropractic, but because neuromusculoskeletal conditions were not in their general sphere of interest. It was noted that if a patient had a complaint unrelated to their specialty, they would normally be referred back to the primary care provider, who would then consult as needed. Sawni and Thomas noted that pediatricians in general practice were more likely than specialists to believe their patients use CAM, to refer for CAM, and to want more CME courses in CAM.15 Borkan et al stated that the physician-patient relationship of physicians in general practice vs. subspecialties may be more open, and physicians in general practice may be more aware of the limitations of biomedicine or deal with less severe, but often chronic, conditions for which CAM may be more appropriate.16
Another interesting finding in the actual referral data is the high referral rates of the women’s health providers, and the OEF/OIF providers. This would suggest a significantly higher rate of female veterans and younger veterans being referred for chiropractic care than would be expected by the relative size of these special populations. This is consistent with anecdotal data concerning the demographics of VA chiropractic clinics. It additionally is very consistent with general chiropractic patient demographics that females utilize chiropractic at a higher rate.17,18
Limitations of this study include sampling error as some providers were reluctant to complete their survey, as they did not want to give their name. This was necessary to compare the self-reported data to actual data to ensure validity of the tool; names will not be required in the future studies. Another limitation would be that many of the specialists do not refer to chiropractors because they practice in specialties that do not treat neuromusculoskeletal (NMS) conditions. The VA utilizes a Primary Care Physician (PCP) gatekeeper model, where patients are referred to specialists by their PCP. In this system chiropractors operate as specialists on a consultant basis. Therefore, future studies may focus more on the referral habits of primary care, and emergency department providers.
Other limitations include the small sample size which does not allow for generalizability of the results of this study. However, the sample size is sufficient for a pilot of the survey instrument. There is also the obvious fact that all providers surveyed practice in a medical center that has a chiropractor on staff. This does not allow a meaningful comparison as there is no control group from a medical center without chiropractors on staff. This data will gain more meaning in future studies when compared to data collected from multiple medical centers with and without chiropractors on staff.
Although there was no control group available in this study for having a chiropractor on staff, the providers self-reported overwhelmingly that having a chiropractor on staff improved their opinion of chiropractic and chiropractors. It was also reported that having a chiropractor on staff increased referrals to chiropractic services.
This study, although limited by a small sample size, accomplishes the primary task of piloting a survey instrument for determining opinions about chiropractic, referral habits concerning chiropractic, demographic characteristics and how all of these are related to one another and to the presence of a chiropractor on staff. Additionally, data was collected and analyzed which can be compared with other VA medical centers to determine the effect having a chiropractor on staff makes toward opinions and referral habits. The results of this study show that providers who know more about chiropractic have a higher opinion of chiropractic, that the satisfaction providers have with chiropractic referral is high, and that the majority of providers who refer to chiropractors do so because they believe it to be the best treatment for that patient. These factors suggest that the presence of a chiropractor on staff will have a positive effect on opinions about chiropractic and therefore on access to chiropractic for patients.
Future studies should evaluate data from hospitals that have staff chiropractors as well as those who don’t. It will be important to evaluate hospitals which are regionally similar and diverse as well as of similar and different sizes. Other factors to consider would be how academically affiliated, and how research oriented the hospital is. Future studies should take advantage of web based surveying applications for easier sampling, and presumably higher penetration with the questionnaires.
This study was supported by the Department of Veterans Affairs. The contents do not represent the views of the Department of Veterans Affairs or the United States Government. The authors would like to thanks Philip Ebrall, PhD for his contributions to design of the study and Melea Fields, EdD for her constant support and tireless editing and manuscript review. Without your support this study could not have been completed.
1. Coulter ID, Singh BB, Riley D, Der-Martirosian C. Interprofessional referral patterns in an integrated medical system. J Manipulative Physiol Ther. 2005;28:170–174.
2. Cherkin D, MacCornack FA, Berg AO. Family Physicians’ views of chiropractors: Hostile or hospitable? American Journal of Public Health.1989 May;79(5):636-7.
3. Verhoef MJ, Sutherland LR. Alternative medicine and general practitioners. Opinions and behaviour. Can Fam Physician. 1995 Jun;41:1005–1011.
4. Easthope G, Tranter B, Gill G. General practitioners’ attitudes toward complementary therapies. Social Science and Medicine. 2000;51:1555–1561.
5. Verhoef MJ, Page SA. Physicians’ perspective on chiropractic treatment. J Can Chiropr Assoc. 1996 Dec; 40(4): 214-219.
6. Sikand A, Laken M. Pediatricians’ experience with attitudes toward complementary/alternative medicine. Arch Pediatr Adolesc Med. 1998;152:1059–1064.
8. Smith M, Greene BR, Haas M, Allareddy V. Intra-professional and inter-professional referral patterns of chiropractors. Chiropr Osteopat. 2006 July; doi:10.1186/1976-1340-14-12.
9. Baker TL. Doing social research. 2nd ed. New York: McGraw-Hill; 1994
10. Langworthy JM, Smink RD. Chiropractic through the eyes of physiotherapists, manual thearapists and osteopaths in the Netherlands. J Altern Complement Med. 2000 Oct;6(5):437-43.
11. Goldszmidt M, Levitt C, Duarte-Franco E, Kaczorowski J. Complementary health care services: a survey of general practitioners’ views. CMAJ. 1995 Jul 1;153(1):29–35.
12. Zotti C, Silvaplana P, Ditommaso S, Russo R, Ruggenini AM. Compulsory and non-compulsory immunizations: Contraindications perceived by medical practitioner. Vaccine. 1992;10(11):742-6.
13. Isberner FR, Simon B, Lloyd LF, Craven JM. PAs and CAM: Clinical experience and recommendations for professional education. Perspective on Physician Assistant Education. 2003;14(3):149-153.
14. Houston EA, Bork CE, Price JH, Jordan TR, Dake JA. How physician assistants use and perceive complementary and alternative medicine. JAAPA. 2001;14:29-30,33-34,39-40.
15. Sawni A, Thomas R. Pediatricians’ attitdues, experience and referral patterns regarding complimentary/alternative medicine: a national survey. BMC Complement Altern Med. 2007 June; doi: 10.1186/1472-6882-7-18.
16. Borkan J, Nehr JO, Anson O, Smoker B. Referrals for alternative therapies. J Fam Pract. 1994 Dec;39(6):545-50.
17. Coulter ID. The chiropractor patient: a social profile. J Can Chiropr Assoc. 1985 March;29(1):25-28.
18. Lishchyna N, Mior S. Demographic and clinical characteristics of new patients presenting to a community teaching clinic. J Chiropr Educ. 2012;26(2):161-168.
Questionnaire for Chiropractic Perception
Please check only one selection unless prompted otherwise. Please pick the selection that best relates to you.
Name__________________________ MD DO PA Specialty _______________________
Gender M F Age 20-34 35-44 45-54 55-64 65-74 75+
Years in practice 0-5 6-10 11-15 16-20 21-30 31+
Does your facility have a chiropractor on staff? Yes No I don’t know
How much do you know about chiropractic?
0 1 2 3 4 5 6 7 8 9 10
Nothing I know everything about it
What is your opinion of chiropractic/chiropractors?
-5 -4 -3 -2 -1 0 1 2 3 4 5
Very negative No opinion Very positive
Has the presence of a chiropractor on the medical staff had an effect on your opinions about chiropractic/chiropractors?
-5 -4 -3 -2 -1 0 1 2 3 4 5
Very negative No effect/ Very positive
How many patients do you see in an average week?
0-14 15-24 25-34 35-44 45-54 55+
What percent of your patients are outpatients?
0 10 20 30 40 50 60 70 80 90 100
How many patients do you consult/refer for chiropractic care per month?
0 1-2 3-4 5-6 7-8 9+
I consult/refer to the chiropractor?
Never When the patient requests it When I believe it to be the best treatment option available
If you have referred to a chiropractor please rate your overall satisfaction on average with the results.
0 1 2 3 4 5 6 7 8 9 10
Not satisfied at all Very Satisfied
In your opinion has the presence of a chiropractor on the medical staff had an effect on your referral patterns?
-5 -4 -3 -2 -1 0 1 2 3 4 5
Refer less No effect/ Refer more
How long is chiropractic training (including any post high school prerequisite study)?
0 1-2 years 3-4 years 5-6 years 7-8 years 8+ years
Which of the following conditions have published research showing chiropractic to be effective in their treatment or management?
Acute low back pain Chronic low back pain Neck pain Headaches Radiating pain Extremity conditions Digestive disorders Hypertension
Infantile colic Fibromyalgia
Technicians/Therapists Mid level providers Independently licensed providers